PDF images at their own resolution

2026-10-29

An image placed in a PDF has a size of its own, counted in pixels, that has nothing to do with how large it looks on the page. A 3000 × 2000 photo can be shown as a small square; a small icon can be stretched across the page. Extract Images saves the image at its stored pixel size — never enlarged, never shrunk to fit the page.

Reading the size

1. Select a card to open the preview.
2. Size shows the stored pixels, for example 2480 × 3508 px.
3. Format shows whether the file will be JPG or PNG; Page shows where it appears.

If a picture looked sharp on screen but the stored size is small, the PDF simply holds a small image. No tool can bring back pixels that were never saved in the file.

Original JPEG data, when possible

Many photos are stored in a PDF as JPEG data. When that data can be used as it is, the image is saved as JPG with the label Original JPEG: the same bytes, not recompressed, so no quality is lost on the way out. Some JPEG data can’t be handed back like that — for example when its colors are CMYK — and the image is saved as PNG instead.

Why other images become PNG

Images stored in other ways are decoded and saved as PNG, which keeps the decoded pixels without adding new loss. The label on each card tells you exactly what happened:

• Colors converted — the image used print colors such as CMYK and was converted to RGB, so print colors may differ.
• Reconstructed — the transparency was rebuilt from the PDF’s mask; edges may differ slightly.
• Mask — a one-color mask: painted areas are black, the rest is transparent.
• Can’t extract — the image can’t be extracted correctly here, so no file is offered rather than a wrong one.

Parts of an image that are cropped away or covered on the page can reappear in the saved file, because the file is the image as stored, not as displayed.